1. Chapter 3 Surveys and Sampling © 2010 Pearson Education 1

2. 2 3.1 Three Ideas of Sampling Idea 1: Examine a Part of the Whole A sample survey is designed to ask questions of a small group of people in the hope of learning something about the entire population. Samples that over- or underemphasize some characteristics of the population are said to be biased. When a sample is biased, the summary characteristics of a sample differ from the corresponding characteristics of the population it is trying to represent. We should select individuals for the sample at random.

3. © 2010 Pearson Education 3 3.1 Three Ideas of Sampling Idea 2: Randomize Randomizing protects us by giving us a representative sample even for effects we were unaware of. Randomization seems fair because nobody can guess the outcome before it happens and because usually some underlying set of outcomes will be equally likely. Sample-to-sample differences are referred to as sampling error.

4. © 2010 Pearson Education 4 3.1 Three Ideas of Sampling Idea 3: The Sample Size Is What Matters The size of the sample determines what we can conclude from the data regardless of the size of the population.

5. © 2010 Pearson Education 5 3.2 A Census—Does It Make Sense? A “sample” that includes the entire population is called a census. A census does not always provide the best possible information about the population: It can be difficult and impractical to complete a census. The population we’re studying may change. Taking a census can be cumbersome.

6. © 2010 Pearson Education 6 3.3 Populations and Parameters Models use mathematics to represent reality. We call the key numbers in those models parameters. A parameter used in a model for a population is called a population parameter.

7. © 2010 Pearson Education 7 3.3 Populations and Parameters Any summary found from the data is a statistic. Sometimes, especially when we match statistics with the parameters they estimate, we use the term sample statistic. A sample that estimates the corresponding parameters accurately is said to be representative.

8. © 2010 Pearson Education 8 3.4 Simple Random Sample (SRS) A sample drawn so that every possible sample of the size we plan to draw has an equal chance of being selected is called a simple random sample, usually abbreviated SRS. With this method each combination of individuals has an equal chance of being selected as well.

9. © 2010 Pearson Education 9 3.4 Simple Random Sample (SRS) A sampling frame is a list of individuals from which the sample will be drawn. Once we have a sampling frame, we can assign a sequential number to each individual in the sampling frame and draw random numbers to identify those to be sampled.

10. © 2010 Pearson Education 10 3.4 Simple Random Sample (SRS) An alternative method is to assign a random number to each member of the sampling frame, sort the random numbers, carrying along the identities of the individuals in the sampling frame, then pick a random sample of any size off the top of the sorted list. Sample-to-sample differences in the values for the variables we measure are called sampling variability.

11. © 2010 Pearson Education 11 3.5 Other Sample Designs Stratified Sampling When we slice the population into homogeneous groups, called strata, use simple random sampling within each stratum, and combine the results at the end, this is called stratified random sampling. Reduced sampling variability is the most important benefit of stratifying.

12. © 2010 Pearson Education 12 3.5 Other Sample Designs Cluster and Multistage Sampling Splitting the population into parts or clusters that each represent the population and performing a census within one or a few clusters at random is called cluster sampling. Sampling schemes that combine several methods are called multistage samples.

13. © 2010 Pearson Education 13 3.5 Other Sample Designs Systematic Samples A systematic sample is created by selecting individuals systematically. To make sure our sample is random, we still must start the systematic selection with a randomly selected individual.

14. © 2010 Pearson Education 14 3.6 Defining the Population Then define your target sample—the individuals for whom you intend to measure responses. Finally, there is your sample—the actual respondents. To start, you should think about the population of interest. Second, you must specify the sampling frame.

15. © 2010 Pearson Education 15 3.7 The Valid Survey A survey that can yield the information you need about the population in which you are interested is a valid survey. To help ensure a valid survey, you need to ask four questions: What do I want to know? Who are the right respondents? What are the right questions? What will be done with the results?

16. © 2010 Pearson Education 16 3.7 The Valid Survey Know what you want to know. You must be clear about what you hope to learn and about whom you hope to learn it. Perhaps the most common error is to ask unnecessary questions.

17. © 2010 Pearson Education 17 3.7 The Valid Survey Use the right sampling frame. A valid survey obtains responses from appropriate respondents. It is important to be sure that your respondents actually know the information you hope to discover.

18. © 2010 Pearson Education 18 3.7 The Valid Survey Ask specific rather than general questions. Watch for biases. If individuals who don’t respond have common characteristics, your sample will suffer from nonresponse bias and will no longer represent the population When respondents volunteer to participate, individuals with the strongest feelings on either side of an issue are more likely to respond; those who don’t care may not bother, creating a voluntary response bias.

19. © 2010 Pearson Education 19 3.7 The Valid Survey Be careful with question phrasing. A respondent may not understand the question the way the researcher intended it.

20. © 2010 Pearson Education 20 3.7 The Valid Survey Be careful with answer phrasing. Inaccurate responses, known as measurement errors, occur when the question does not take into account all possible answers. The best way to prevent measurement errors is a pilot test, in which a small sample is drawn from the sampling frame, and a draft form of the survey instrument is administered.

21. © 2010 Pearson Education 21 What Can Go Wrong? In a voluntary response sample, a large group of individuals is invited to respond, and all who do respond are counted. Voluntary response samples are almost always biased, and so conclusions drawn from them are almost always wrong. In convenience sampling we simply include the individuals who are convenient. Unfortunately, this group may not be representative of the population. An SRS from an incomplete sampling frame introduces bias because the individuals included may differ from the ones not in the frame.

22. © 2010 Pearson Education 22 What Can Go Wrong? Many survey designs suffer from undercoverage, in which some portion of the population is not sampled at all or has a smaller representation in the sample than it has in the population. Rather than sending out a large number of surveys for which the response rate will be low, it is often better to design a smaller, randomized survey for which you have the resources to ensure a high response rate.

23. © 2010 Pearson Education 23 What Can Go Wrong? Surveys that are too long are more likely to be refused, reducing the response rate and biasing all the results. Response bias includes the tendency of respondents to tailor their responses to please the interviewer and the consequences of slanted question wording.

24. © 2010 Pearson Education 24 What Can Go Wrong? How to Think about Biases Look for biases in any survey. Spend your time and resources reducing biases. If you possibly can, pretest or pilot your survey. Always report your sampling methods in detail.

25. © 2010 Pearson Education 25 What Have We Learned? A simple random sample (SRS) is our standard. Stratified samples can reduce sampling variability. Cluster samples randomly select among heterogeneous subgroups that each resemble the population at large. Systematic samples can work in some situations. Multistage samples combine several random sampling methods.

26. © 2010 Pearson Education 26 What Have We Learned? Nonresponse bias can arise when sampled individuals will not or cannot respond. Response bias arises when respondents’ answers might be affected by external influences. Voluntary response samples are almost always biased and should be avoided.

27. © 2010 Pearson Education 27 What Have We Learned? Convenience samples are likely to be flawed. Even with a reasonable design, sample frames may not be representative. Look for biases in any survey. We’ve learned to report our methods so that others can evaluate the fairness and accuracy of our results.